arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.09649cs.LGstat.ML
收藏

多目标多武器土匪的约束遗憾最小化

Constrained regret minimization for multi-criterion multi-armed bandits

Stanford University(斯坦福大学);IIT Bombay(印度理工学院孟买分校);IIT Madras(印度理工学院马德拉斯分校)

Anmol Kagrecha, Jayakrishnan Nair, Krishna Jagannathan

摘要:

考虑随机多臂强盗环境,研究了在给定时间范围内的约束后悔最小化问题。每个手臂都与一个未知的,可能是多维的分布相关联,手臂的优点是由几个可能相互冲突的属性决定的。其目的是优化一个“主”属性受用户提供的其他“次要”属性的约束。我们假设属性可以使用来自臂分布的样本来估计,并且估计量具有适当的集中性质。我们提出了一种称为Con-LCB的算法,它保证了对数遗憾,即,所有非最优臂的平均游戏次数在视界中至多是对数。该算法还输出布尔标志,该布尔标志以高概率正确地识别给定实例相对于约束是否可行/不可行。我们还表明,Con-LCB是最佳的一个普遍的常数,即,更复杂的算法不可能普遍地做得更好。最后,我们建立了一个基本的权衡后悔最小化和可行性识别。我们的框架发现自然的应用程序,例如,在金融投资组合优化,风险约束最大化的预期收益是有意义的。

英文摘要:

We consider a stochastic multi-armed bandit setting and study the problem of constrained regret minimization over a given time horizon. Each arm is associated with an unknown, possibly multi-dimensional distribution, and the merit of an arm is determined by several, possibly conflicting attributes. The aim is to optimize a 'primary' attribute subject to user-provided constraints on other 'secondary' attributes. We assume that the attributes can be estimated using samples from the arms' distributions, and that the estimators enjoy suitable concentration properties. We propose an algorithm called Con-LCB that guarantees a logarithmic regret, i.e., the average number of plays of all non-optimal arms is at most logarithmic in the horizon. The algorithm also outputs a Boolean flag that correctly identifies, with high probability, whether the given instance is feasible/infeasible with respect to the constraints. We also show that Con-LCB is optimal within a universal constant, i.e., that more sophisticated algorithms cannot do much better universally. Finally, we establish a fundamental trade-off between regret minimization and feasibility identification. Our framework finds natural applications, for instance, in financial portfolio optimization, where risk constrained maximization of expected return is meaningful.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑